Research Journal of Applied Sciences, Engineering and Technology 6(6): 1106-1110, 2013
ISSN: 2040-7459; e-ISSN: 2040-7467
© Maxwell Scientific Organization, 2013
Submitted: November 08, 2012
Accepted: December 28, 2012
Published: June 30, 2013
Research on Evaluation of Component-Based Software Testing
Xueyun Ji and Zihui Che
Department of Mathematics and Computer, Baoding University, Baoding 071000, Hebei, China
Abstract: The research object of our work is evaluating program for the component-based software testing. As an
important part into the software engineering, software testing is the primary instrumentality to guarantee the quality
and reliability of the software. With the Component-Based Software Development (CBSD) method becoming a
popular method of software engineering, people pay more and more attention to the quality of component software.
Therefore, the component-based software testing is an important method to guarantee the quality of component
software. There are many factors existing in the component-based software testing process, in order to avoid some
errors and defects and improve the testing process, we need evaluate the validity of component-based software
testing. Based on the discussion of the quality characteristics of the software component, the study studies the
workflow and the relevant measures which apply to evaluate the validity of component-based software testing.
Keywords: Component, component-based software testing, validity of software testing, quality, workflow
INTRODUCTION
Software quality (Chen, 2008) is the consistent level
that between the software product and the needs which
defined explicitly and implicitly. Software quality
evaluation national standard ISO 14958 defines software
quality as “the combination of software properties, the
ability that software can meet the users’ demands which
specified explicitly or potential”. Software testing is the
assurance of software quality. Software testing (Hou and
Cui, 2008) can not only find out errors that existing in
the software, but also can help finding the defects during
the software development process, through analyzing the
causes and stages that make errors, to improve the
software development process and the software quality.
With component technology become mature and the
CBSD (Hu and OuYang, 2008) method uses widely,
software developers enjoy the advantages which bring
by component technology, such as development rapidly,
higher reusability and so on. And at the same time, they
pay more attention to the quality of component software
and the problem of component software testing. There
are many studies about component software testing now,
researchers have proposed many feasible methods and
testing strategy, such as the testing method based on
component metadata (Ma et al., 2006) , the testing
method based on UML (Unified Modeling Language,
UML) (Shang and Zhang, 2006), the testing method
based on interface mutation (Andres and Macario, 2009)
and so on. Software testing occupy a large proportion in
the process of software development, the percentage of
workload of testing in the total workload is usually
between thirty and forty percent. In order to improve the
testing process and use the developing cost effectively, it
is necessary to evaluate the validity of software testing.
At present, in the field of studies about the evaluation of
software testing, there are many research results, for
example, the method of evaluation of software testing
quality combined with fuzzy theory (Yao et al., 2007),
the dynamic evaluation method based on multi
dimensional test coverage (An et al., 2010) and an
evaluation model of validity of software testing based on
Bayesian theorem (Wang, 2007).
The study studies the component-base software
testing and studies the evaluation of the validity of
component software testing to ensure the quality of
component software.
THE QUALITY OF SOFTWARE COMPONENT
Component is the smallest unit in the component
software system; its quality influences the quality of
component software directly. The industry pays more
and more attention to the importance of the quality of
software component. For software developers, they
concern about evaluating the quality of software
component to improve the quality of software
component product; for the person who reuse software
component, they are more worried about how to
evaluate the quality of software component to select
suitable software component from a mass of component
library.
Software component has the quality characteristics
which are also suitable for the generally software.
Corresponding Author: Xueyun Ji, Department of Mathematics and Computer, Baoding University, Baoding 071000, Hebei,
China
1106
Res. J. Appl. Sci. Eng. Technol., 6(6): 1106-1110, 2013
Fig. 1: The model of external quality and internal quality of software component
Fig. 2: The model of using quality of software component
Generally speaking, the quality model which is used to
describe software product can also apply to component.
According to the basic quality properties of general
software, combined with the characteristics of software
component, we can get the quality model of software
component. Similar to the general software, the quality
model of software component also divides three parts;
they are external quality model, internal quality model
and using quality model. We get the external and
internal quality model through the following process: on
the basis of the basic quality properties in the software,
combining with the component characteristics, we add a
quality property which is related to component, which is
reusability, as shown in Fig. 1.
Using quality is the quality from users’ view, its
acquiring is dependent on obtaining the required external
quality and the external quality is dependent on
obtaining the required internal quality. Form users’
view, if the component can be reuse or not and the times
of reusing reflect the trust degree of users use the
component directly. The using quality model of software
component is as shown in Fig. 2. We can see from
Fig. 2, on the basis of the original using quality
properties of software, it increases a using quality
property, which is the reliability of software component.
The reliability has two factors: reusing degree and
choosing degree. Reusing degree identifies the reusing
reliability of software component and choosing degree
identifies the expected selection of software component,
it shows a proportion that potential user will use this
component.
Fig. 3: The classification of measuring the validity of testing
THE MEASUREMENT OF
VALIDITY OF TESTING
The development of component and component
software bring along the study and development of
component software testing. The main purpose of
evaluating the validity of testing is to evaluate testers’
work and use the evaluation results to improve testing
process. During component software testing process, it
usually has some invalid aspects; the goals of evaluation
include identifying invalid problems in order to take
repair measures. The classification of common means
which measure the validity of testing is as shown in
Fig. 3.
The measurement of customers’ satisfaction is a
method that measures afterwards, this method can’t get
the measurement results before the product is in sales,
installation and using.
The common measurement method of validity of
testing is build around defect analysis. Defect analysis is
the following series of activities (Xin, 2009). It does
necessary collection with defects which are generated in
the process of software development and running
(Agrestl and Evanco, 1992). Then classify the
1107 Res. J. Appl. Sci. Eng. Technol., 6(6): 1106-1110, 2013
information about the defects, do summary statistics,
calculate the analytic indicator and compile the
analytical reports. Defect analysis provides some
indexes of software reliability (Aroul and Lavergne,
1996). The indexes provide the judging standards to
reveal the trend or the distribution of defects. The
common defect analysis indexes include the detection
rate of defect, the density of defect, the elimination rate
of defect and the loss of defect and so on:
D 
Dn
Fn
(1)
We define (1) is the computing formula for the
density of defect, which Dρ denotes the density of
defect, Dn denotes the number of defects and Fn denotes
the number of code lines or function points.
DRE 
Dtn
Dtn  Ndn
(2)
We define (2) is the computing formula for the
elimination rate of defect, which DRE expresses the
elimination rate of defect, Dtn expresses the number of
defect that are found during testing process and Ndn
expresses the number of defect that are undiscovered.
DL 
Dn  Fg
Tdn
(3)
We define (3) is the computing formula for the loss
of defect, which DL means the loss of defect, Fg means
the weight values of preclinical stages that find out
defect, Tdn means the total number of defects.
The measurement of coverage is a practical
measurement method of the validity of testing. The
coverage can measure the testing sets or the integrity of
actual executing testing.
THE WORKFLOW AND THE RELVANT
MEASURES OF EVALUATING COMPONENT
SOFTWARE TESTING
The workflow of evaluating the validity of
component software testing: During the process of
evaluating the validity of testing, we need to focus on
two aspects: the aim of evaluation and the required
information categories that can achieve the aim of
evaluation. Making right aim can guide the whole
evaluating process and the collection of accurate
information can ensure the evaluating process
smoothly. Figure 4 shows the workflow of evaluating
the validity of testing.
The evaluating work of validity is begin when the
testing process is finished, its input information are the
ending executing results of testing stage, these input
information include the following types: the number of
executed testing, the resources that are consumed in
testing, the tool that are used in testing, the defect that
are discovered, the size of software that is tested, the
number of days that is used to repairing defect, the
defect that are not repaired, the defect are found in
operation that are undiscovered in testing, the phases
that find out defect and the name of discovered defect
and so on. The executing process of evaluating the
validity contains seven aspects: that are making the aim
of evaluation, determining the contents of evaluation,
making the responsibility of measurement, choosing the
method of evaluation, determining the required facts,
collecting the data of evaluation and evaluating the
validity of testing. After the whole executing process,
according to the quality checklists, we can inspect the
evaluating process, part quality inspection items are as
shown in Table 1. Then the reasonable evaluation can
output the improved testing process, otherwise it need to
return to the start of the executing process, remake the
aim of evaluating and execute the evaluating process
again.
Fig. 4: The evaluating flow of the validity of software testing
1108 Res. J. Appl. Sci. Eng. Technol., 6(6): 1106-1110, 2013
Table 1: A part of checking item of quality
Serial number
The description of project
Project 1
Whether the management department support the continuous improving the process of testing or not
Project 2
If the resources that used to improve the process of testing are available
Project 3
If the results of testing accumulate with time growing
Project 4
If the result of testing include the various of component items which are described in input parts
Project 5
If the analysis result is reasonable
Project 6
If the executing analysis is based on the characteristic of component software testing
Project 7
If the analysis result is used in improving the testing process
Table 2: The measures to evaluate the testing projcet of component software
Serial
number
Measures
Using measures
1
The time cost in testing
The degree of testing
2
The personnel involved in The degree of testing
testing
3
The changing rate of the The degree of testing
schedule of testing
4
The number of testing
The degree of testing
5
The component that is tested The degree of testing
6
The cost of testing process
7
The cost that locate defect
8
The budget that achieved
The deviation rate of the actual testing schedule
compared to the schedule in planning
The number of testing and the scale of the tested system
The number of tested component and the total number of
component
The resources that consumed during The cost of testing and the total cost of the whole system
testing
The resources that consumed during The cost of testing and the number of defect that is
testing
located during testing process
The resources that consumed during The expected cost of testing and the actual cost of testing
testing
Table 3: The measures to evaluate the testing process of component software
Serial number Measures
Using Measures
1
The errors that are found in product
The validity of testing
2
The defect that are found during testing
The validity of testing
3
The defect of interface that are found during
testing
The validity of business by testing
The times that the component reusing
The validity of the unit testing of component
The validity of testing
4
5
6
7
8
The description of measures
The time cost during testing process
The number of people that participate in testing
The validity of the integration testing of
component
The Validity Of The Testing Stage
The validity of testing
Involving system
The validity of testing
The validity of testing
The Validity Of Testing
The measures of evaluating component software
testing: We need a measuring standard to do objective
evaluation for the validity of testing program, by
analyzing the measures, the evaluators can adopt
targeted measures, record the results of evaluation, so it
is necessary to establish indexes system to evaluating
component software testing. The traditional software
testing programs have three aspects of measurement:
the measurement of project, the measurement of
process and the measurement of product. Combining
with the traditional measures, this study establishes the
measures to evaluate the testing of component software
system. Table 2 and 3 show a part of measures.
The description of measures
The number of errors that are found in product and the scale of the
system
The defect that is located during testing and the total defect of the whole
system
The defect of the interface of component that are assigned by testing
The loss cost by problem and the total resources of the system
The times of the component was reused
The cost of unit testing and the number of errors are found during this
process
The cost of integration testing and the number of errors are found during
this process
The cost of testing and the number of errors are found during debugging
which we proposes to evaluate the validity of component
software testing, it can improve the process of
component-based software testing, play the role of
component-based software testing bitterly and improve
the quality of component-based software.
REFERENCES
Agrestl, W.W. and W.M. Evanco, 1992. Projecting
software defects from analyzing software designs.
IEEE T. Softw. Eng., 18: 988-997.
An, J., G. Wang, S. Liu and J. Zhu, 2010. Dynamic
evaluation method based multi-dimensional test
coverage for software testing. J. Softw., 21(9):
2135-2147.
Andres, F. and P. Macario, 2009. Testing-based process
CONCLUSION
for evaluating component replace ability. Electr.
Notes Theor. Comput. Sci., 236(2): 101-115.
From component software testing, the study
Aroul, M.E. and C. Lavergne, 1996. Generalized linear
discusses external quality model, internal quality model
models in software reliability parametric & semiand using quality model of software component, on the
paremetric approaches. IEEE T. Rel., 45(3):
research of the measurement of validity of testing, it
463-470.
proposes the workflow to evaluate the validity of
Chen, P., 2008. Evalvation of software quality. Softw.
component-based software testing and at the time it
Guide, 7(1): 32-34.
gives the related measuring indexes. Using the workflow
1109 Res. J. Appl. Sci. Eng. Technol., 6(6): 1106-1110, 2013
Hou, H. and D. Cui, 2008. Research on assurance of
software quality based on software testing
technology. Sci. Technol. Inform., 13: 58.
Hu, G. and S. OuYang, 2008. Componet-based
software product development and investigation.
Control Eng. China, 15(1): 75-81.
Ma, L., Y. Lu and M. Liu, 2006. Application research
of metadata on generating component-based
software test case. Mini-Micro Syst., 27(12):
2228-2231.
Shang, X. and Y. Zhang, 2006. Research of UMLbased generating test case for component
integration testing. Comput. Eng., 32(1): 96-98.
Wang, D., 2007. An evaluation model of validity of
software testing based on Bayesian theorem. Mod.
Bus. Trade Ind., 19(11): 268-269.
Xin, F., 2009. The reason and the analysis for the bug
appeared in software. J. Hebei Acad. Sci.,
2009(26): 28-30.
Yao, Y., X. Liu and S. Huang, 2007. An assessment
method for fuzzy partial ordering relation-based
software testing quality. J. Univ. Electr. Sci. Tech.
China, 36(3): 503-509.
1110